{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 32
        },
        "id": "HMwI-ZC4vKYs",
        "outputId": "908a010a-c86e-4246-96e9-f62ac4e389a9"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {}
        }
      ],
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "\n",
        "fig = plt.figure()\n",
        "\n",
        "x = [200, 300, 400]\n",
        "y1 = [0.3938396176314392 , 0.4302708443972385, 0.44503451938396166]\n",
        "yerr1 = [0.014272184414012389, 0.018305733848378577, 0.011139764717686148]\n",
        "y2 = [0.5348911311736589, 0.5472118959107807, 0.5595326606479023]\n",
        "yerr2 = [0.018559724542714137, 0.015179984895792472 , 0.016832743779634632] "
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "fig, (ax1, ax2) = plt.subplots(1, 2)\n",
        "fig.set_size_inches(9,4)\n",
        "plt.subplots_adjust(left=0.1, bottom=0.15, right=0.95, top=0.9, wspace=0.2, hspace=0.15)\n",
        "\n",
        "ax1.axhline(y=0.41, color='#42f578', linestyle='-.', lw=3)\n",
        "(_, caps, _) = ax1.errorbar(x, y1, yerr1, capsize=20, elinewidth=10, linewidth=0, color='#4251f5')\n",
        "for cap in caps:\n",
        "    cap.set_color('red')\n",
        "    cap.set_markeredgewidth(10)\n",
        "ax1.set_xlim(150, 450)\n",
        "ax1.set_title(\"44-Topic Classification\")\n",
        "ax1.set_xlabel(\"Number of Training Samples\")\n",
        "ax1.set_ylabel(\"Top-1 Accuracy\")\n",
        "ax1.set_xticks(x)\n",
        "\n",
        "ax2.axhline(y=0.507, color='#42f578', linestyle='-.', lw=3)\n",
        "(_, caps, _) = ax2.errorbar(x, y2, yerr2, capsize=20, elinewidth=10, linewidth=0, color='#4251f5')\n",
        "for cap in caps:\n",
        "    cap.set_color('red')\n",
        "    cap.set_markeredgewidth(10)\n",
        "ax2.set_xlim(150, 450)\n",
        "ax2.set_title(\"8-Topic Classification\")\n",
        "ax2.set_xlabel(\"Number of Training Samples\")\n",
        "ax2.set_xticks(x)\n",
        "\n",
        "plt.savefig(\"Figure_1.pdf\")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 292
        },
        "id": "eTwypAcRxILx",
        "outputId": "6d0c8f3b-3bc8-4105-8153-91eac51f25cd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 648x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}